Development and Characterization of an Ethyl Methane Sulfonate (EMS) Induced Mutant Population in Capsicum annuum L.

Plant breeding explores genetic diversity in useful traits to develop new, high-yielding, and improved cultivars. Ethyl methane sulfonate (EMS) is a chemical widely used to induce mutations at loci that regulate economically essential traits. Additionally, it can knock out genes, facilitating efforts to elucidate gene functions through the analysis of mutant phenotypes. Here, we developed a mutant population using the small and pungent ornamental Capsicum annuum pepper “Micro-Pep”. This accession is particularly suitable for mutation studies and molecular research due to its compact growth habit and small size. We treated 9500 seeds with 1.3% EMS and harvested 3996 M2 lines. We then selected 1300 (32.5%) independent M2 families and evaluated their phenotypes over four years. The mutants displayed phenotypic variations in plant growth, habit, leaf color and shape, and flower and fruit morphology. An experiment to optimize Targeting Induced Local Lesions IN Genomes (TILLING) in pepper detected nine EMS-induced mutations in the eIF4E gene. The M2 families developed here exhibited broad phenotypic variation and should be valuable genetic resources for functional gene analysis in pepper molecular breeding programs using reverse genetics tools, including TILLING.


Introduction
Pepper (Capsicum annuum L.) is an economically important crop worldwide. Peppers are consumed as food for their richness in nutritional and medicinal substances, including capsaicinoids, carotenoids, and vitamins A and C. They are also used in the cosmetic and pharmaceutical industries, and for ornamental purposes [1,2]. Breeding of modern pepper cultivars focuses on improving economically important traits, including yield and nutrient and secondary metabolite content [3][4][5][6]. These programs require vast genetic diversity, which can be promoted by mutagenesis to alter nucleotide sequences, potentially creating novel alleles useful for crop improvement [7,8]. There are four main categories of agents used to alter the genome: physical mutagens (e.g., radiation with X-rays, gamma-rays, fast neutrons, or UV); chemical agents (e.g., ethyl methane sulfonate (EMS), ethyl nitrosourea (ENU), 1,2:3,4-diepoxybutane (DEB), and N-nitroso-N-methylurea (NMU); biological agents (e.g., T-DNA and

Effect of EMS dose on Seed Germination
Micro-Pep is a decorative pepper plant with short internodes, whorled phyllotaxy, lanceolate leaves, and attenuate rounded fruits. Its compact growth and small size make the accession suitable for mutation studies. The EMS concentration used in the treatment of seeds is crucial for the successful development of a mutant population. Therefore, we tested the effect of different EMS concentrations (1, 1.3, 1.5, and 2% EMS) on germination two weeks after placing the seeds in moist petri dishes. As expected, the percentage of germinated seeds decreased with the increase in EMS concentration ( Figure 1). The lowest germination (59%, compared with 96% in the control) was obtained with a treatment with 2% EMS, and the highest (91%) with 1% EMS (Figure 1). The germination markedly decreased from 80 to 69% with 1.3 and 1.5% of EMS, respectively ( Figure 1). Based on these results, to ensure that we obtain a sufficient population after mutagenesis, we selected the 1.3% concentration to generate a mutant pepper population.

Phenotypes of M 1 Mutant Plants
Of the 9500 seeds treated with EMS, 6620 (70%) germinated on medium (Table 2). We transplanted 4210 seedlings to pots with soil, scored their phenotypes, and harvested the seeds of the 3996 that  (Table 2). The M 1 seedlings included mutants with aberrant phenotypes for plant height, foliage color, and inflorescence development (Table 3). In terms of plant height, 45 and 18 plants were scored as tall (>17 cm) and short (<10 cm) mutants, respectively ( Table 3). The average height of normal Micro-Pep was 12.5 cm. We observed chimeras in the M 1 population with albino or variegated foliage. Overall, 23 mutant individuals had foliage discoloration (Table 3). Flowerless mutants were the most abundant mutant category in the M 1 generation, with 231 individuals displaying only vegetative growth (Table 3). Male sterility was also observed: 41 individuals had male-sterile flowers (Table 3). This abundance of mutant phenotypes confirmed that mutagenesis was successful and implied that a high number of mutants would be recovered in the subsequent M 2 generation.

Characterization of M 2 Mutant Phenotypes
Of the 1300 M 2 lines selected for phenotypic analysis, 372 showed mutant phenotypes, which corresponded to a mutation frequency of 28.6% (Table 4). We divided the phenotypic alterations into four main classes: plant growth and habit, leaf color and morphology, flower characteristics, and fruit color and morphology ( Figure 2 and Table 4). These main mutant categories included 11 subclasses ( Figure 2). Mutants were named according to their phenotype category and compared with the wild-type Micro-Pep plants ( Figure 3). Among the mutants, 30.5% had variations in plant growth and habit, 50.2% in leaf color and morphology, 12.3% in flower characteristics, and 7% in fruit color and morphology ( Figure 2 and Table 4). Most of the mutations showed recessive inheritance, and many plants had pleiotropic phenotypes, in which individuals were altered in more than one trait (e.g., flowerless and dwarf individuals).  confirmed that mutagenesis was successful and implied that a high number of mutants would be recovered in the subsequent M2 generation.

Characterization of M2 Mutant Phenotypes
Of the 1300 M2 lines selected for phenotypic analysis, 372 showed mutant phenotypes, which corresponded to a mutation frequency of 28.6% (Table 4). We divided the phenotypic alterations into four main classes: plant growth and habit, leaf color and morphology, flower characteristics, and fruit color and morphology ( Figure 2 and Table 4). These main mutant categories included 11 subclasses ( Figure 2). Mutants were named according to their phenotype category and compared with the wild-type Micro-Pep plants ( Figure 3). Among the mutants, 30.5% had variations in plant growth and habit, 50.2% in leaf color and morphology, 12.3% in flower characteristics, and 7% in fruit color and morphology ( Figure 2 and Table 4). Most of the mutations showed recessive inheritance, and many plants had pleiotropic phenotypes, in which individuals were altered in more than one trait (e.g., flowerless and dwarf individuals).

Plant Growth and Habit Phenotypes
The most conspicuous mutant class observed in the M 2 generation was plant growth and habit ( Figure 4). The mutant phenotypes included in this category were: aberrant plant height (tall or dwarf when compared with wild-type plants) ( Figure 4A,B); retarded growth with no visible stem ( Figure 4C); and abnormal branching with a multi-whorled canopy ( Figure 4D). Of the mutant lines in the plant growth and habit, 37 were dwarves, 7 were tall, 15 had retarded growth, and 48 had abnormal branching (Table 4). These mutants can be used to explore the genetic mechanism underlying plant height in pepper breeding programs.

Flower Color and Morphology Phenotypes
We observed three categories of mutants related to flower characteristics: inflorescence, organ color, and morphology (Table 4). Inflorescence mutants included 3 with twin flowers, 27 flowerless,

Flower Color and Morphology Phenotypes
We observed three categories of mutants related to flower characteristics: inflorescence, organ color, and morphology (Table 4) Table 4). A unique mutation related to flower organ color rendered the anthers pink ( Figure 7D; Table 4). M 2 lines with altered flower morphology included one filamentous-stamen, four AGAMOUS-type, one fasciculation-type, three with swelled ovary with short stamen, and one shell mutant ( Figure 7E-I; Table 4). These flower mutants can be used to study the genes involved in pepper flower morphology and organ development.  Table 4). A unique mutation related to flower organ color rendered the anthers pink ( Figure 7D; Table 4). M2 lines with altered flower morphology included one filamentous-stamen, four AGAMOUS-type, one fasciculation-type, three with swelled ovary with short stamen, and one shell mutant ( Figure 7E-I; Table 4). These flower mutants can be used to study the genes involved in pepper flower morphology and organ development.

Fruit Color and Morphology
Several different types of mutations affected fruit morphology and color ( Figure 8; Table 4). For morphology, we separated mutants according to alterations in fruit shape: round fruit, cylindrical shape type, oval shape, multi-fruit on single calyx, pyramid shape, and two fruit shapes on one plant. Those were present in 5, 2, 3, 1, 4, and 1 mutant lines, respectively ( Figure 8A-F; Table 4). In contrast to the wild-type Micro-Pep fruits, which are red at maturity, seven mutant lines carried

Fruit Color and Morphology
Several different types of mutations affected fruit morphology and color ( Figure 8; Table 4). For morphology, we separated mutants according to alterations in fruit shape: round fruit, cylindrical shape type, oval shape, multi-fruit on single calyx, pyramid shape, and two fruit shapes on one plant. Those were present in 5, 2, 3, 1, 4, and 1 mutant lines, respectively (Figure 8A-F; Table 4). In contrast to the wild-type Micro-Pep fruits, which are red at maturity, seven mutant lines carried orange fruits, and one line had yellow and red fruits on the same plant ( Figure 8G-H; Table 4). These mutants related to fruit shape represent potential breeding material to alter pepper fruit shape, as well as being resources to help elucidate the underlying genetic mechanisms.

Mutation Screening and Detection by TILLING
We next examined the suitability of TILLING analysis for mutant pepper populations, and tested this method to the eIF4E gene of Micro-Pep and Yuwol-cho M2 lines. This analysis revealed nine putative mutations (Figure 9; Table 5), which were confirmed by Sanger amplicon sequencing ( Table 5). The eIF4E gene structure in pepper consists of five exons and four introns ( Figure 9A). Two of the mutations were in Micro-Pep pooled samples, whereas seven were in Yuwol-cho (Table 5; Figure 9B). Five of the mutations localized to 'Exon1', three were detected in 'Exon2-3', and one was in 'Exon 4-5' (Figure 9B; Table 5). We confirmed the presence of single nucleotide mutations and their positions at the base-pair level ( Table 5). The confirmed mutations all represented G:C to A:T base-pair transitions. These results confirmed that TILLING is a suitable method to find mutations in important pepper genes. Hence, the population developed in this study can serve in the future as an inclusive platform for reverse genetics studies in pepper.

Mutation Screening and Detection by TILLING
We next examined the suitability of TILLING analysis for mutant pepper populations, and tested this method to the eIF4E gene of Micro-Pep and Yuwol-cho M 2 lines. This analysis revealed nine putative mutations (Figure 9; Table 5), which were confirmed by Sanger amplicon sequencing ( Table 5). The eIF4E gene structure in pepper consists of five exons and four introns ( Figure 9A). Two of the mutations were in Micro-Pep pooled samples, whereas seven were in Yuwol-cho (Table 5; Figure 9B). Five of the mutations localized to 'Exon1', three were detected in 'Exon2-3', and one was in 'Exon 4-5' ( Figure 9B; Table 5). We confirmed the presence of single nucleotide mutations and their positions at the base-pair level ( Table 5). The confirmed mutations all represented G:C to A:T base-pair transitions. These results confirmed that TILLING is a suitable method to find mutations in important pepper genes. Hence, the population developed in this study can serve in the future as an inclusive platform for reverse genetics studies in pepper.

Discussion
Mutagenesis by EMS is an effective approach to create genetic diversity in plant populations. In this study, we treated the dwarf Capsicum annuum L. accession Micro-Pep with 1.3% EMS and characterized the phenotypes of the M1, M2, and M3 generations. We analyzed 1300 M2 families and observed copious mutant phenotypes. A total of 372 families (28.6%) had clear mutations that affected the important traits of plant height, plant habit, leaf color, leaf morphology, and flower and fruit-related features.

Discussion
Mutagenesis by EMS is an effective approach to create genetic diversity in plant populations. In this study, we treated the dwarf Capsicum annuum L. accession Micro-Pep with 1.3% EMS and characterized the phenotypes of the M 1 , M 2 , and M 3 generations. We analyzed 1300 M 2 families and observed copious mutant phenotypes. A total of 372 families (28.6%) had clear mutations that affected the important traits of plant height, plant habit, leaf color, leaf morphology, and flower and fruit-related features.
Genetic diversity is fundamental to the success of breeding programs. Hence, it is imperative to maximize the effectiveness and efficiency of the mutagens used to generate that diversity [8,21]. The effect of a mutagen depends both on the concentration applied and on the germplasm treated. Therefore, it is necessary to optimize the procedure to assure a high mutation frequency without compromising seed viability [8,33]. This optimization is essential for EMS because a high concentration drastically reduces seed germination in multiple species [13]. The optimum dosage of EMS for rice, soybean, and tomato is below 1% [13,21,34] but it is higher (1.5%) for some pepper cultivars [7,35]. Here, we started by testing the effect of different EMS concentrations on Micro-Pep germination. Then, we treated bulk seeds with 1.3% EMS, which allowed 80% of germination and produced a mutation frequency high enough to create multiple mutants with observable phenotypes.
Mutant phenotypes are less likely to emerge in the M 1 generation, when only dominant mutations can be identified [36]. We identified several mutant phenotypes in M 1 plants, including seedling growth defects, changes in plant height, alteration in foliage color, absence of inflorescences, and sterility. Previous mutation studies in pepper and eggplant also reported mutant phenotypes in M 1 plants [17,18], though some of those were not present in subsequent generations. [35] and [21] similarly reported that many of the mutants observed in M 1 pepper seeds mutagenized with EMS were not identified in the next generations. The mutation frequencies observed in the M 1 generation may fluctuate substantially in M 2 or M 3 due to the non-heritability of large deletions that occur during mutagenesis [12]. M 2 families of mutagenized populations reveal the recessive mutations and can show marked trait variation [17,21]. In accordance, the M 2 generation of our study displayed striking mutant phenotypes with variable mutation frequencies. As seen in previous studies [17,37], plant height-related mutations were the most frequent. Dwarf mutants are fundamental for the elucidation of the regulatory mechanisms behind plant growth and development. They are also crucial for breeding of lodging-resistant cultivars [8]. We identified several dwarf mutants with abnormal branching, shorter internodes, or retarded growth with no obvious stems. Other dwarf mutants in pepper [4,7,21,38] are likely caused by the suppression of epidermal cell expansion or defects in gibberellin (GA) biosynthesis [39]. Identification of new recessive, monogenic mutations causing dwarfism could lead to the discovery of novel genes responsible for pepper growth and inform modern breeding programs [40].
We also identified chlorophyll mutants among the M 2 lines, with a range of albino, yellow, pale-green, and dark-green phenotypes. The presence of chlorophyll mutants is a good indicator of the effectiveness of a mutagen in pepper [14], and EMS-treated pepper populations often display chlorotic and whitened leaves [7,21]. Here, the chlorotic mutants showed a range of mottled yellow and pigmented patterns. Mutants with leaf-color alterations are valuable for discovering genes responsible for chlorophyll metabolism [41]. Additionally, can be used as indicators of seed purity in the breeding of new cultivars in ornamental plants with better photosynthetic efficiency [42]. The leaf-color mutants identified here can provide alternative foliage color choices for breeders of ornamental peppers.
We also observed mutants altered in leaf morphology, causing long petioles, and scabrous or proliferate leaves. In addition, we isolated several mutants with upward or downward curled leaves. Leaf curling can improve light energy absorption and photosynthetic rate. In addition, it can modulate leaf transpiration and enhance drought tolerance. Therefore, curled-leaf mutants are appealing genetic resources for breeding drought-tolerant cultivars [43]. Notably, the curling phenotype was present throughout the full growth period, irrespective of the moisture and temperature, unlike the recessive pepper flc mutant in which leaves are flat at night or under appropriate moisture and curled during the day [19]. This difference suggests that we have isolated novel alleles that affect leaf curling which is required further study to confirm.
The M 2 population also included mutant lines with defects in inflorescence development or without reproductive growth, suggesting that the genes controlling flower induction might be mutated in these individuals. The AGAMOUS mutants are well characterized flower mutants in Arabidopsis [44], and we identified agamous-like mutants with inheritance stability. We selfed those M 2 lines and observed a 3:7 segregation ratio of mutant to wild type in the M 3 generation (data not shown). These mutant lines will be useful resources in flower biology and organ development-related research. We also identified several mutant lines that were male sterile. Male sterility is often induced by mutagenesis and is a desirable trait for breeding hybrids [45,46]. However, cytoplasmic male sterility is induced only rarely in crops such as sugar beet and pearl millet [47,48]. Test crosses and allelism assays will be needed to define the type of male sterility obtained here.
Fruit shape is critical for the market value of a horticultural commodity. Mutants with changes in fruit shape can lead to breeding profitable new cultivars according to market demand. We characterized several mutants with round, oval, cylindrical, and pyramid-shaped fruits. Fruit shape is challenging to measure, and it is a quantitative trait in nature. Additionally, the mechanisms controlling the fruit orientation, size, and shape remain elusive in pepper [4,22]. The mutants reported here will be useful for genetic studies aimed at uncovering the mechanisms that govern fruit shape-related traits. Fruit color is another essential trait for the value of commercial pepper cultivars. We identified a line with orange fruits that had stable inheritance, and a recent study employed this mutant to investigate fruit color variation in pepper [49]. Thus, the mutants identified here can serve as a primary material for identifying and characterizing the genes that regulate fruit color and morphology.
Many mutant studies in crops involve chemical mutagenesis-based TILLING experiments. Chemical mutagenesis, including EMS, generally produce single point mutations. Mutation densities among diploids have varied from 1 mutation per 150 kb to over 1 mutation per 1 million bp [50,51]. In our TILLING experiment, we analyzed the eIF4E gene with different primer sets. We discovered nine induced point mutations across the sets of chemically mutagenized populations. EMS converts GC to AT due to recurrent alkylation of guanine remnants [52]. In major plant model species, including Arabidopsis, wheat, maize, and pea, more than 99% of the mutations detected were GC:AT transitions [53]. We also detected point mutations that led to GC:AT base conversions here. Recently, TILLING methods are undergoing a series of adjustments to allow incorporation with high-throughput next-generation sequencing (NGS) technologies. Advanced NGS platforms like Illumina, SOL-iD, Ion Torrent, and PacBio and the steady reduction in sequencing costs has also contributed to the combination of NGS and TILLING methods. This approach can now be used efficiently in grains and horticultural crops [54][55][56]. Hence, the mutant population developed here can be analyzed by TILLING and sequencing, allowing the fast and robust detection of mutant alleles for use in pepper breeding in the near future.

Conclusions
Here, we employed EMS to generate genetic variability in a dwarf and compact pepper germplasm, Micro-Pep. We observed and characterized M 1 , M 2 , and M 3 mutant phenotypes in plant growth and habit, leaf color and morphology, flower characteristics, and fruit color and morphology. Two of the phenotypes showed stable inheritance up to the M 3 generation. With TILLING, we discovered nine putative point mutations in the eIF4E gene, four of which were confirmed by sequencing. Our Micro-Pep EMS mutant population showed great genetic diversity, making it a valuable resource for reverse genetic studies. These include TILLING approaches to determine the genetic factors underlying the phenotypes observed. In addition, availability of the pepper genome will facilitate map-based cloning of mutations of interest. Furthermore, newly developed strategies based on whole-genome sequencing can also be exploited. We have submitted our phenotypic data for the mutant population to an online browser (https://phenome-networks.com/) that is freely accessible to the scientific community. Finally, the mutants will be exchanged and shared among pepper breeders and researchers.